Bacterial biofilms predominate in both acute and chronic human lung infections

Abstract

Background: A basic paradigm of human infection is that acute bacterial disease is caused by fast growing planktonic bacteria while chronic infections are caused by slow-growing, aggregated bacteria, a phenomenon known as a biofilm. For lung infections, this paradigm has been thought to be supported by observations of how bacteria proliferate in well-established growth media in the laboratory-the gold standard of microbiology.

Objective: To investigate the bacterial architecture in sputum from patients with acute and chronic lung infections.

Methods: Advanced imaging technology was used for quantification and direct comparison of infection types on fresh sputum samples, thereby directly testing the acute versus chronic paradigm.

Results: In this study, we compared the bacterial lifestyle (planktonic or biofilm), growth rate and inflammatory response of bacteria in freshly collected sputum (n=43) from patient groups presenting with acute or chronic lung infections. We found that both acute and chronic lung infections are dominated by biofilms (aggregates of bacteria within an extracellular matrix), although planktonic cells were observed in both sample types. Bacteria grew faster in sputum from acute infections, but these fast-growing bacteria were enriched in biofilms similar to the architecture thought to be reserved for chronic infections. Cellular inflammation in the lungs was also similar across patient groups, but systemic inflammatory markers were only elevated in acute infections.

Conclusions: Our findings indicate that the current paradigm of equating planktonic with acute and biofilm with chronic infection needs to be revisited as the difference lies primarily in metabolic rates, not bacterial architecture.

Keywords: bacterial infection; cystic fibrosis; pneumonia; respiratory infection.

Figure 1

Flow chart of study and final diagnoses of patients recruited. The majority of patients screened and subsequently not recruited did not meet the requirements for untreated acute infection. CAP, community-acquired pneumonia; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease.

Figure 2

Biofilms and planktonic cells are observed across infection types. (A–C) Representative projections of confocal images of sputum samples of CAP with no detected pathogen (A), COPD with Moraxella sp (B) and CF with Pseudomonas aeruginosa (C). Specimens were stained with Tamra-5 (red) using PNA-FISH probes specific to bacterial 16S rRNA and DAPI m(blue). Scale bar is 10 µm. (D) Median total biomass of bacteria by infection type. Bacterial biomass was calculated on each sample (n=43) by counting the voxels representing bacteria after image analysis pipeline. There was no significant difference in sample biomass between infection types (p<0.05, Kruskal-Wallis test). (E) Comparing median sample intensity across infection type. Bacterial objects on each sample were identified, classified as either planktonic cells (≤5 µm³) or biofilms (>5 µm³) and their per cent contribution to total biomass was calculated. We found CAP samples to have the higher median intensity than CF samples (p<0.05, Kruskal-Wallis test), while COPD and CF samples, typically described chronic infections, have equivalent median intensity. (F) Comparing median sample intensity in biofilm (red) vs planktonic cells (black). We also found the median intensity of voxels in biofilms is higher than in planktonic cells (p<0.0001, Wilcoxon test). CAP, community-acquired pneumonia; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease; DAPI, 4′,6-diamidino-2-phenylindole; PNA-FISH, peptide nucleic acid fluorescent in situ hybridisation.

Figure 3

Estimation of bacterial growth rate based on fluorescence intensity. (A) The median intensity of bacterial voxels in biofilm (red) and planktonic (black) cells from infection types. We calculated the intensity emitted by biofilm and planktonic cells from each infection type and compared CAP infection samples with both COPD and CF samples. We found that CAP samples have higher intensity in both biofilm and planktonic cells than COPD (p=0.0356, Kruskal-Wallis test), suggesting that CAP infections have an increased growth rate compared with COPD and CF. (B) Per cent of bacterial population at maximum intensity. We calculated the portion of voxels at the highest fluorescence intensity value of biofilm (red) and planktonic (black) cells on each sample. We compared biofilms to planktonic cells within each infection type and found that biofilms have higher fluorescence intensity that planktonic cells in all infection types (p<0.01, unpaired Wilcoxon test). CAP, community-acquired pneumonia; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease.

Figure 4

Host cell biomass in all infection types. (A) Mean total biomass of inflammatory cells by infection type. Confocal images of sectioned sputum were stained with DAPI and biomass calculated by the total number of and blue fluorescent voxels. (B) Distance from bacteria at which the proportional occupancy of inflammatory cells is highest for each sample type. (C) Proportional occupancy of inflammatory cells relative to bacteria. Representative samples of each type are shown: CAP (left), COPD (middle) and CF (right). Each point is the average value from 1000 random voxels in the image. (D) Blinded histopathological evaluation of sputum samples of degree of inflammation from sputum samples: from CAP (n=16), COPD (n=11) and CF (n=14). Degree of inflammation: 0: no inflammation, 1: mild inflammation, 2: moderate inflammation and 3: severe inflammation. Statistical significance was determined using (A) and (B) ordinary one-way ANOVA followed by Bonferroni multiple comparison test and (D) Kruskal-Wallis test (p≤0.05). ANOVA, analysis of variance; CAP, community-acquired pneumonia; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease; DAPI, 4’,6-diamidino-2-phenylindole